| opendevreview | Clay Gerrard proposed openstack/swift master: Include % free when FALLOCATE_RESERVE fails https://review.opendev.org/c/openstack/swift/+/904870 | 02:21 |
|---|---|---|
| opendevreview | Matthew Oliver proposed openstack/swift master: Make the dark data watcher work with sharded containers https://review.opendev.org/c/openstack/swift/+/787656 | 04:12 |
| mattoliver | ^ just added a note about the sharding misplaced object edgecase. I hope that's enough that we can land the patch now :) | 04:14 |
| opendevreview | Jianjian Huo proposed openstack/swift master: Proxy: Use namespaces when getting listing/updating shards https://review.opendev.org/c/openstack/swift/+/895602 | 05:41 |
| opendevreview | Matthew Oliver proposed openstack/swift master: wip: shard replication sync points https://review.opendev.org/c/openstack/swift/+/905064 | 07:49 |
| mattoliver | ^ that is just a WIP, it's completely untested for the moment. Just getting it off my computer. | 07:50 |
| acoles | zaitcev: nice, thanks for sharing the slides | 08:59 |
| opendevreview | Merged openstack/swift master: Make the dark data watcher work with sharded containers https://review.opendev.org/c/openstack/swift/+/787656 | 09:47 |
| opendevreview | Alistair Coles proposed openstack/swift master: ContainerBroker.get_shard_ranges(): states must be a list https://review.opendev.org/c/openstack/swift/+/904861 | 12:19 |
| opendevreview | Alistair Coles proposed openstack/swift master: Container-server: add container namespaces GET https://review.opendev.org/c/openstack/swift/+/890470 | 12:19 |
| opendevreview | Alistair Coles proposed openstack/swift master: Proxy: Use namespaces when getting listing/updating shards https://review.opendev.org/c/openstack/swift/+/895602 | 12:19 |
| opendevreview | Alistair Coles proposed openstack/swift master: proxy-server: only fetch all namespaces when memcache exists https://review.opendev.org/c/openstack/swift/+/905112 | 12:42 |
| zaitcev | acoles: If you know something that makes us stand out in Python sense, please let me know. There's not much time, it's tomorrow, but I'd love to add a slide. | 14:00 |
| k-s | guys, quick q: https://opendev.org/openstack/swift/src/branch/master/swift/common/middleware/acl.py#L229-L244 parse_acl() will call the v1/2 handlers... but the result is not converted. The v1 is a tuple, while v2 is a dict. It may also return None, yet all of the users (keystoneauth.py, tempauth.py, s3api/controllers/acl.py) and even doc (source/development_auth.rst) assumes it's always a tuple | 14:08 |
| k-s | is it a bug? would you take a patch to change those to always be a tuple? | 14:08 |
| opendevreview | Alistair Coles proposed openstack/swift master: sq: container server test fixups https://review.opendev.org/c/openstack/swift/+/905124 | 15:19 |
| clayg | k-s it's probably silly that's it's a single function - the original author may have planned on eventually providing a backwards compatible migration for container-acls from v1 to v2 and that work was never done. | 15:52 |
| clayg | k-s: right now container-acls are most widly supported and always v1; only tempauth and some out-of-tree-proprietary auth systems https://platform.swiftstack.com/docs/cookbooks/swift_usage/account_acl.html support account-level ACL and those are always v2 | 15:54 |
| clayg | zaitcev: how did the talk go? I assume more people would be interested in asyncio+ASGI than eventlet+WSGI; but IME generator based concurrency + blocking io in python thread pool is just not as fast as evenlet/greenlet/stack-swapping + small blocking ios! | 16:04 |
| zaitcev | clayg: it's tomorrow, I'm soliciting improvements | 16:04 |
| clayg | when jianjian tried ASGI+aiosqlite the results were not encouraging - it's made me unexcited about ASGI+aiofiles - but I have hope for more "native" asyncio tools like io_uring! https://github.com/qweeze/uring_file | 16:05 |
| opendevreview | Alistair Coles proposed openstack/swift master: sq: container server test fixups https://review.opendev.org/c/openstack/swift/+/905124 | 16:14 |
| opendevreview | Alistair Coles proposed openstack/swift master: proxy-server: only fetch all namespaces when memcache exists https://review.opendev.org/c/openstack/swift/+/905112 | 16:14 |
| opendevreview | Tim Burke proposed openstack/swift master: s3api: Refactor how we handle expected errors https://review.opendev.org/c/openstack/swift/+/751555 | 18:29 |
| timburke | oh man, ^^ reminds me that i had p 613452 going a while back... | 18:35 |
| patch-bot | https://review.opendev.org/c/openstack/swift/+/613452 - swift - s3api: Move authenticator logic to separate module - 4 patch sets | 18:35 |
| opendevreview | Alistair Coles proposed openstack/swift master: Alternative: Include % free when FALLOCATE_RESERVE fails https://review.opendev.org/c/openstack/swift/+/905143 | 19:07 |
| opendevreview | Alistair Coles proposed openstack/swift master: Alternative: Include % free when FALLOCATE_RESERVE fails https://review.opendev.org/c/openstack/swift/+/905143 | 19:15 |
| acoles | patch-bot ❤️ | 19:20 |
| timburke | it's been a while, thought i'd bring it back finally :-) | 19:27 |
| opendevreview | Merged openstack/swift master: ContainerBroker.get_shard_ranges(): states must be a list https://review.opendev.org/c/openstack/swift/+/904861 | 20:38 |
| opendevreview | Jianjian Huo proposed openstack/swift master: Container-server: add container namespaces GET https://review.opendev.org/c/openstack/swift/+/890470 | 20:40 |
| opendevreview | Jianjian Huo proposed openstack/swift master: Proxy: Use namespaces when getting listing/updating shards https://review.opendev.org/c/openstack/swift/+/895602 | 20:44 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!